Column: Analysts shouldn’t sweat poor track record with U.S. corn, soy yields

4 minute read

Soybean fields are inspected as part of University of Wisconsin research trial into whether the weed killer dicamba drifted away from where it was sprayed in Arlington, Wisconsin, U.S., August 2, 2018. REUTERS/Tom Polansek/File Photo

Register now for FREE unlimited access to

NAPERVILLE, Ill., Aug 9 (Reuters) - Grain analysts have struggled to anticipate the U.S. government’s corn and soybean yields in past Augusts, but that does not necessarily mean their instincts about actual crop potential were wrong.

The U.S. Department of Agriculture’s August crop report, due on Friday at noon EDT (1600 GMT), is notorious for its surprising nature. It has been nine years since USDA’s August corn and soybean yields both fell within the pre-report range of analyst guesses.

But stacked up against final yields, the government seems to have just as many problems in August.

Register now for FREE unlimited access to

In percentage terms, the average August trade guess for corn yield was closer to the final than USDA’s August forecast in 13 of the last 23 years, and the trade was closer on soybeans in 14 years.

That trend for corn is not as heavy in more recent years than it was earlier this century, but the trade was closer than USDA in August in three of the last four years, excluding 2019. The trade was closer on soybeans the last three years, but USDA was better in August in the previous four.

Actual magnitudes do not vary much in corn. In absolute percentage terms, the trade’s last five August corn yields deviated an average of 2.6% from the final versus USDA at 2.7%.

The trade has been distinctly closer on soybeans in August with yields an average of 1.7% from the final compared with USDA at 2.3%. That suggests analysts’ poor track record may not exactly be deserved, they are just bad at guessing what USDA’s method will produce this month.

Last year is a good example of the trade bombing in August despite solid intuition. Analysts had corn yield at 177.6 bushels per acre and USDA shocked to the low side with 174.6, but final came in at a record 177.


USDA’s corn yield has landed outside the trade range in six of the last eight Augusts, including 2021, and for soybeans that has happened in five of the last seven Augusts. In all five of those soybean cases, USDA’s yield came in unexpectedly high.

Friday’s trade estimates are not safeguarding against a range miss in corn. The average trade guess of 175.9 bpa sits in a 4.4 bpa range from low to high, above last year’s 4.3 but below the five-year average of 6.2.

That yield would be down from USDA’s trend of 177 bpa, the fourth-highest yield on record but the third-highest printed in August. Trend yield is computed by USDA’s World Board and the August figure is put together by USDA’s agricultural statistics service (NASS).

The trade is set on a below-trend corn yield, another risk in addition to the narrow estimate range. Of the 24 analysts polled by Reuters, just one pegged yield above 177 bpa while 21 offered a lower number.

Within the last eight years, the August corn yield has come in below the July trend twice: 2017 and 2021. One commonality 2022 shares with those years is relatively lower early August conditions. U.S. corn was 58% good-to-excellent as of Sunday versus 60% in the same week in 2017 and 64% in 2021.

USDA’s August corn yield came in over the trade range in 2017 but under it in 2021.

NASS in 2019 eliminated field sampling in its August report, instead putting emphasis on producer surveys and satellite data. It is too early to tell if this move has affected the agency’s August accuracy, but it is something to consider when comparing past outcomes.

U.S. August corn yield: USDA versus analysts


Analysts predict U.S. soybean yield at 51.1 bpa, with a low of 49.9 and a high of 52.5. That range of 2.6 bpa is in line with the five-year average.

However, only one of 24 estimates is above USDA’s trend of 51.5 bpa. Eighteen call for something lower.

Last year was the first since 2014 where August soybean yield was below the trade average, and the same is true for corn yield. However, the trade’s average soy underestimation of 1.75 bpa in the previous six years caused some painful days for bulls.

Last year was the first since 2013 where USDA’s August soy yield came in lower than the trend. Crop conditions do not seem to explain that as 60% good-to-excellent in early August last year was associated with a different yield direction than the same score in 2017.

Soybeans were 59% good or excellent as of Sunday, which aside from 2019 is the lowest score for the week since 2012. That same stat is true of this week’s corn conditions.

U.S. August soybean yield: USDA versus analysts

Karen Braun is a market analyst for Reuters. Views expressed above are her own.

Register now for FREE unlimited access to
Writing by Karen Braun Editing by Matthew Lewis

Our Standards: The Thomson Reuters Trust Principles.

Opinions expressed are those of the author. They do not reflect the views of Reuters News, which, under the Trust Principles, is committed to integrity, independence, and freedom from bias.

Thomson Reuters

As a columnist for Reuters, Karen focuses on all aspects of the global agriculture markets with a primary focus in grains and oilseeds. Karen comes from a strong science background and has a passion for data, statistics, and charts, and she uses them to add context to whatever hot topic is driving the markets. Karen holds degrees in meteorology and sometimes features that expertise in her columns. Follow her on Twitter @kannbwx for her market insights.